A New Generalized Deep Learning Framework Combining Sparse Autoencoder and Taguchi Method for Novel Data Classification and Processing

Table 2

Autoencoder 1 upper and lower level values of factors for DDOS detection.

Factors	Lower Limit	Upper Limit
Hidden Size (HS)	19	22
Max Epochs (ME)	300	500
L2 Weight Regularization (L2)	0.0035	0.0045
Sparsity Regularization (SR)	4	6
Sparsity Proportion (SP)	0.13	0.16